Provides guidance to statistical offices on their possible roles in developing and producing leading, composite and sentiment indicators, and gives practical and operational guidance to statistical offices that produce or consider producing these indicators. Existing examples of good practice are also referenced.